In a conventional wireless packet communication method, a radio channel to be used is determined in advance. Prior to transmission of data packets, carrier sense is performed to detect whether or not that radio channel is idle. Only when that radio channel is idle, one data packet is transmitted. This management process enables a plurality of STAs to share one radio channel in a staggered manner ((1) International Standard ISO/IEC 8802-11 ANSI/EEE Std. 802.11, 1999 edition, Information technology—Telecommunications and information exchange between systems—local and metropolitan area networks—Specific requirements—part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications; (2) “Low-powered Data Communication System/Broadband Mobile Access Communication System (CSMA) Standard, ARIB STD-T71 version 1.0, Association of Radio Industries and Businesses, settled in 2000).
In order to improve a maximum throughput of the above wireless packet communication method, a method is known which broadens the frequency band per radio channel so as to increase the data transmission rate in the PHY layer.
However, it is necessary to provide a transmission deferral duration of a predetermined length that is independent of the data transmission rate in the PHY layer immediately after the transmission of a packet in order to avoid collision of packets, as pointed out, for example, in an article (Iizuka et al., “5 GHz Wireless LAN System Based on the IEEE802.11a standard—Packet Transmission Characteristics—”, B-5-124, Proceedings of the Electronics Information and Communication Engineers, Society Conference 2000, September 2000). When such a transmission deferral duration is provided, the transmission efficiency of data packets (a ratio of the maximum throughput to the data transmission rate in the PHY layer) decreases with the increase in the data transmission rate in the PHY layer. Thus, it is difficult to significantly improve the throughput by only increasing the data transmission rate in the PHY layer.
On the other hand, the application of a MIMO technique (Kurosaki et al., “100 Mbit/s SDM-COFDM over MIMO Channel for Broadband Mobile Communications”, Technical Reports of the Institute of Electronics, Information and Communication Engineers, A•P 2001-96, RCS2001-135(2001-10)) has been considered for improving the maximum throughput without broadening the frequency band per radio channel. In the MIMO technique, different data packets are transmitted from a plurality of antennas on the same radio channel at the same time. The data packets transmitted at the same time on the same radio channel are received by digital signal processing that can deal with the difference in propagation coefficients of the respective data packets received by a plurality of wireless antennas in an opposed STA. A MIMO number is determined in accordance with the propagation coefficients and the like.
On the other hand, in the case where each STA has a plurality of wireless network interfaces and can use a plurality of radio channels, different radio channels can be used between a plurality of STAs. In this case, improvement of the throughput is promising, as compared with communication performed by time-dividing one radio channel.
However, when the center frequencies of a plurality of radio channels used at the same time are close to each other, the effect of leakage power, which leaks from one radio channel to a frequency band used by another radio channel, becomes large. In general, in the case of transmitting a data packet, after a transmit-side STA sends a data packet, a receive-side STA sends back an acknowledge packet (ACK packet, NACK packet) for the received data packet to the transmit-side STA. When the transmit-side STA tries to receive that acknowledge packet, the effect of leakage power from another radio channel that carries out transmission at the same time becomes a problem.
For example, the case is considered where the center frequencies of radio channels #1 and #2 are close to each other and the transmission time is different between data packets that are transmitted simultaneously from the radio channels #1 and #2, as shown in FIG. 28. In the shown example, the data packet transmitted from the radio channel #1 is shorter. Thus, when an ACK packet for that data packet is received, the data packet on the radio channel #2 is still in transit. Therefore, the radio channel #1 may not receive the ACK packet because of leakage power from the radio channel #2. In this situation, the throughput cannot be improved even if transmission is carried out by using a plurality of radio channels at the same time.
If transmission rates of the respective radio channels are the same, the above situation occurs because of a difference in packet time length (transmission time=data size) between data packets. Considering the transmission rates of the respective radio channels, the above situation occurs because of the difference in packet time length (transmission time=data size/transmission rate) between data packets.
In a wireless LAN system or the like, the data size of a data frame input from a network is not constant. Thus, in the case where input data frames are sequentially converted into data packets and are transmitted, the packet time length (transmission time) of each data packet also changes. Therefore, even when a plurality of data packets are transmitted at the same time, as shown in FIG. 28, it is highly likely that a packet time length difference occurs between the data packets and causes a failure in the receiving of the ACK packet.
Moreover, in a wireless LAN system operating in accordance with IEEE802.11 Standard, for example, a data frame input from a wired network (e.g., Ethernet (registered trademark) frame) is converted into a MAC (Media Access Control) frame and a data packet generated from that MAC frame is transmitted as a wireless packet to a wireless line.
In a conventional system, one data frame is converted into one MAC frame from which one data packet is generated. Therefore, even if the data size of a data field in a data frame is small, that data frame is converted into one MAC frame and is transmitted as one data packet (wireless packet). For example, the maximum size of the data field of a MAC frame in accordance with IEEE802.11 Standard is 2296 bytes, whereas the data size of the data field of an Ethernet (registered trademark) frame that is typically used as a data frame is limited to 1500 bytes. Therefore, even if the Ethernet (trademark) frame has the maximum size, that Ethernet frame has some leftover in the maximum size (2296 bytes) of the data field of the MAC frame. That is, the conventional system does not effectively use the maximum data size that can be transmitted in one MAC frame, and has a limitation in the improvement of the throughput.
As described above, in order to further improve the throughput, it is necessary to overcome the problem of different packet time lengths in the case of simultaneous transmission using a plurality of radio channels (i.e., different data sizes when the transmission rates of the radio channels are the same) and the problem of inefficiency in the case where the data size of a data frame is smaller than the maximum size of a data field of a MAC frame.
FIG. 29 shows an exemplary configuration of a wireless LAN system. In FIG. 29, a mobile terminal (address: S1) 11, a server (address: S2) 12, and a server (address: S3) 13 are connected through an access point (hereinafter AP) (address: AP) 10. The mobile terminal 11 and the AP 10 are connected to each other through a wireless line on which wireless packets are transmitted. The AP 10 and the servers 12 and 13 are connected through a router and the Internet. Ethernet (registered trademark) frames are transmitted between them. In this example, a transmission direction from the servers 12 and 13 to the mobile terminal 11 is called as a “downlink”, and a transmission direction from the mobile terminal 11 to the servers 12 and 13 is called as an “uplink”.
FIG. 30 shows a frame format for the downlink of the wireless LAN system. In FIG. 30, an Ethernet frame transmitted from the server 12 or 13 to the AP 10 contains a header, a frame body, and FCS. In the header of the Ethernet frame transmitted from the server 12 to the mobile terminal 11, a destination address DA and a source address SA are set to “S1” and “S2”, respectively. The frame body accommodates an IP packet for which the destination address DA and the source address SA are set to “S1” and “S2”, respectively.
The Ethernet frame transmitted from the server 12 or 13 is converted into a wireless packet in the AP 10 and is transmitted to the mobile terminal 11. The wireless packet contains a MAC header, a frame body, FCS, and the like. In the MAC header of the wireless packet corresponding to the Ethernet frame transmitted from the server 12, “S1” is set as a destination address DA in a PHY layer, “AP” is set as BSS (Basic service set) ID corresponding to a source address in the PHY layer, and “S2” is set as a source address SA. The frame body accommodates an IP packet for which “S1” is set as the destination address DA and “S2” is set as the source address SA.
FIG. 31 shows a frame format for the uplink of the wireless LAN system. In FIG. 31, the mobile terminal 11 generates an IP packet to the server 12 or 13. In the IP header of the IP packet to the server 12, “S2” is set as a destination address DA and “S1” is set as a source address SA. A wireless packet transmitted from the mobile terminal 11 to the AP 10 contains a MAC header, a frame body, FCS, and the like. In the MAC header of the wireless packet accommodating the IP packet to the server 12, BSSID corresponding to a destination address in the PHY layer is set to “AP”, a source address SA in the PHY layer is set to “S1”, and a destination address DA is set to “S2”. The frame body accommodates the IP packet for which the destination address DA is set to “S2” and the source address SA is set to “S1”.
The wireless packet transmitted from the mobile terminal 11 is converted into an Ethernet frame in the AP 10 and is transmitted to the server 12 or 13. The Ethernet frame contains a header, a frame body, and FCS. In the header of the Ethernet frame to the server 12, a destination address DA and a source address SA are set to “S2” and “S1”, respectively. The frame body accommodates the IP packet for which the destination address DA and the source address SA are set to “S2” and “S1”, respectively.
A conventional AP generates one wireless packet from one Ethernet frame to the mobile terminal and transmits that wireless packet on the downlink of the PHY layer from the AP to the mobile terminal. From a viewpoint of improving transmission efficiency, a method is considered to be effective that generates one or more wireless packets from a plurality of Ethernet frames transmitted from a plurality of servers to the same mobile terminal and then transmits those one or more wireless packets in one lump.
Similarly, a conventional mobile terminal generates one wireless packet from one IP packet to a server and transmits that wireless packet on the uplink of the PHY layer from the mobile terminal to the AP. From a viewpoint of improving transmission efficiency, a method is considered to be effective that generates one or more wireless packets from a plurality of IP packets to the same server and transmits those one or more wireless packets in one lump.
Moreover, for the uplink of the PHY layer from the mobile terminal to the AP, another method can be considered, in which one or more wireless packets are generated from not only IP packets to the same servers but also IP packets that are transmitted to different servers through the same AP, and are transmitted in one lump. In the mobile terminal, IP packets to the same server can be found by referring to their destination addresses DA. It is also possible to find whether or not a plurality of IP packets to different servers go through the same AP by checking their destination addresses DA against the AP address.
It is an object of the present invention to provide a wireless packet communication method that can easily generate a plurality of data packets having the same packet time length in the case of transmitting a plurality of data packets simultaneously between two STAs by using a plurality of radio channel at the same time. It is another object of the present invention to provide a wireless packet communication method that can generate a plurality of data packets from one Ethernet frame or IP packet or from a plurality of Ethernet frames or IP packets to the same destination for each of a downlink and an uplink in a PHY layer between a mobile terminal and an access point (hereinafter AP) or between STAs.